Is there a win32 ext2/3 filesystem driver out there anywhere?
Forget that -- there is FAT code in the Linux kernel. More IP that smacks Linux and means that it cannot be distributed (and interoperate with windows, as FAT-based systems were the only major filesystem that both Linux and Windows can read and write out-of-box. Very bad juju.
FWIW, it is *damned* hard to write Windows filesystem drivers -- compare a small Linux filesystem -- RAMFS, at 342 lines of source -- with even a minimal Windows driver. There is an ext2 implementation with read support, though.
Oh, yes. The embedded community uses FAT all over the place. They are going to absolutely go bonkers when this hits the news.
I know that it's not quite what you wanted, but using Coda, which is designed to support disconnected operation (i.e. servers goes away for a while and then comes back) may be an appealing option for you.
Does that make the facts in his post any less relevant or true?
That particular element does not, he's also made plenty of errors in his posts.
However, I do claim that if he wants his posts to be uncriticized, he should remove the false credentials and try for recognition based on the value of the content in them. He has made the choice to impersonate someone else -- he has to live with people criticizing him.
Furthermore, there is social damage caused by impersonating other people to the Slashdot community and propagating false information. This is why people that make errors in their posts frequently see corrections posted. It is the intelligent thing to do for other Slashdot members to resist impersonation becoming accepted.
Sigh...but I can't reasonably expect a Linux version.
Damn it, I should really sit down and spend a couple weeks trying to figure out what would be involved in some sort of system that would allow binary distribution of (fast) software under Linux that wouldn't break in twenty four months. Stallman may love open source, but the world's always going to have some form of binary-only software...and Linux trails Windows in this arena terribly.:-(
It was one of the first games to leave Blizzard-style micromanagement. The interface is designed so that using it isn't one of the challenges of the game (limited group sizes, queues, etc), but to help you as much as possible (flexible AI toggles per group or unit, easy to queue up masses of units and preassign orders, etc).
It was the first I know of to have really neat sea battles. Infantry were cheap -- you could churn out tons -- but each ship was *expensive*, and specialized. The first time I played a sea battle level, I was enthralled.
It had great explosions and fires.
Battles took place over more realistic ranges -- people didn't shoot the equivalent of twenty feet. The biggest guns could lob rounds from seven or more screens away.
There was no limit on resources. A round didn't come to an end because you exhausted your resources -- you used everything possible, just as intelligently as you could.
There were masses of intelligent auto-build and repair abilities.
And a ton of other things.
Cavedog (TA's publisher) could have gone far, but for two factors: Blizzard, it's main competitor, didn't make as good games but had a phenomenal marketing budget that it used well, and TA's sequel, TA:Kingdoms, really sucked compared to TA.
Incidently, the guy that designed the TA system (where you could tell things to follow things that attacked them, or not etc)...I believe his name was "Tim" something...went on to make some medieval game with the same style interface. It wasn't an RTS, though. I can't remember the name. Fantastic to see that one game designer is interested in making a highly usable interface, not one that you have to fight.
I've read both aritcles and I'm still confused. What the heck is the point and why should I care?
Brief summary (and I'm not informed about this particular spat). LinuxGazette was originally a newsletter that one guy decided to put out to "help people have a just a little more fun" with Linux. He reminded me a lot of Cliff Stoll -- folksy, nice, and very into just helping people out.
As it happened, more and more volunteers started joining, and the LG became one of the early sources for good Linux information and tips.
The folks that publish Linux Journal (a commercial, nonvolunteer publication) started getting involved at some point, paying for an editor and hosting.
Now, the publisher (which has put money and time into Linux Gazette) and at least a large chunk of the volunteers disagree about whether or not to change the format of the gazette.
Frankly, I really don't see the problem. I understand that if you've spent a lot of time working on something, you get upset if you lose it.
Here's my take:
SSC/Linux Journal should get the trademark (legally, if not morally). They registered the thing. It's ugly, but that's life. Not sure whether the fact that it was around for a long time pre-registration is an issue.
SSC/Linux Journal needs to stop modifying copyright information, removing article authors and replacing them with itself. That was a complaint of the forkers. They are quite in the legal right.
The readers deserve to be informed. Putting up information about this on both sites and linking to each other is important.
The readers will go to whatever they want -- if one sucks, they'll leave it.
It's sad that such friendly volunteer work wound up in such a legal tangle. I guess that the moral of the story is: avoid companies unless you're damned sure that they won't screw you over.
When did this fixation that books were somehow 'superior' to visual media first come into vogue? I've seen some very moving movies in my time, and read some awful books.
Sure (for a *very* fun combination, try reading Do Cyborgs Dream of Electric Sheep? and watching Blade Runner). However, the LoTR series is a very, very good series of books. It's a good set of movies, but just good, not very, very good. It's as if someone was a master oil painter and someone made a good watercolor reproduction of one of their works -- the watercolor just isn't a substitute for the original.
Either Mirriam-Webster is in error, or it is referring to a highly unusual and archaic grammar usage.
Keep in mind that Mirriam-Webster may be considered authoritative (well, *I* consider it authoritative, though the really anal-retentive folks will look for the OED) for spelling -- not for grammar. I have always seen English style references state that use of the apostrophe-"s" combination is incorrect. All of the serious resources that Google digs up on short notice (aside from m-w) also back this -- take a look at the MLA style apostrophe guidelines or this linguist's lengthy analysis with an eye on "BOUNDARY MARKER (EXCLUSIVELY)".
Just to ensure that people are aware of the fact -- Samir Gupta is not a PhD, does not work for Nintendo, and is one of Slashdot's more colorful frauds. You can see the beginning of the Samir Gupta hoax on USENET years ago, in early discussions on the Sega Genesis.
However, an amazing number of new people with mod points, impressed with the bogus credentials, frequently mod up his posts.
I'm starting to suspect that the United States will never see conversion.
We have fast enough and cheap enough hardware now that it's feasible (nicer) home connections to stream down much-better-than-TV video over an Internet connection. There are a number of improvements to make in upstream distribution structure, but ultimately, despite the fact that IP currently provides essentially nothing by way of real-time guarantees, my guess is that we'll slowly start seeing more and more Internet-based systems. It just doesn't make sense to have a single purpose dedicated system just for TV.
I suspect that those cheap consumer broadband routers will start having a "smart bandwidth allocation" feature that the ISP will also grok which guarantees real-time delivery (well, over the last and slowest leg of the trip). It wouldn't be a very difficult system to devise -- system on local network allocates bandwidth from router, router talks to upstream system.
A healthy amount of precaching would be important -- this could be an issue in sports, where having a sub-one-minute precache is essential to many hardcore fans. It'd work wonderfully for almost anything else, though.
It'd only really be practical if you put together a massive collection of.torrents (like the suprnova database) in a.zip file and then used BT to download that.
Good explanation, but it still wouldn't work. (This is an interesting problem that you'll run into in a lot of areas of computer science, though I don't know whether it has a name.) Basically, each.torrent points to a file. The contents of that file may not be changed without invalidating the.torrent link. Thus, the contents of the file must be known at the time of propagating the.torrent. Thus, the file may only reference older.torrents -- that is,.torrents that were already in existence at the time the file was created. So if the only reference someone has to obtain BitTorrent information is a.torrent file, they may only ever reach.torrents created at or before the creation time of that.torrent.
Your solution would reduce load over the existing system, but from the fact that the parent mentioned chicken/egg, I suspect that they want to use BitTorrent as a complete solution.
There are systems that allow providing forward references -- allowing one to reach newer files. These systems require the use of encryption, not just hashing, however. kast and freenet both use public key cryptography to provide exactly this service.
Oh, and if you think you can steal market shares from, let us say, Sun, without them making a fuss, I think you are mistaken too. Last time I checked, Sun is still worth more money than Red Hat...
Last time I checked, Solaris was losing market share rapidly to Linux. Dunno how much of that is to Red Hat Linux, but we can surmise a fair amount.
If Red Hat isn't marketing a UNIX clone, then what's it marketing now? Last time I checked, Linux is a UNIX clone. Sure, it's not SCO UNIX(R)(TM), but it's still UNIX. Sometimes I wonder whether these MBAs really know what the hell they're trying to sell or if they just have a form process to market anything.
No, what he said was exactly right.
"We are making a product foo, which is a clone of bar. Foo competes mostly with bar, and will kill off bar within a decade."
How hard is that to understand?
Weavers are a clone of triscuits, and saying that "triscuits will be dead within the decade, killed by weavers" is an entirely valid statement.
Soon we may wind up with something where the proverbial "Joe Sixpack" pays relatively high fees on his Skype phone he bought at Wal-mart and plugged into the wall, while all the "techies" pay nothing to use their "alternative" VoIP setups.
Remember "audio" CD-Rs? It can happen, little sense as it makes...
MMO is waaay overblown. Simple reason. There just aren't enough players. MMOGs take time to play. There just aren't enough people willing to devote a significant enough chunk of their life to a game. There are some, and they were enough to keep Ultima Online flush with players. There were even enough for Everquest as well. But there are now MMOGs coming out at an incredible clip, which keeps fragmenting the customers base further, and reducing the value of other competitors in the field. MMOGs are flopping left and right, and still publishers frantic for a pirate-resistant subscription model try to start new MMOGs.
How much money has GTK+ made for GNU? How much money has LGPL wxWindows made? How about plain XFree86? I'm not talking about donations from Redhat or SuSE, I'm talking about actual revenue from actual customers. Now ask yourself if that's enough to support even one full time developer?
Yes, but instead of looking at it from a TrollTech point of view, look at it from a user point of view. Regardless of *why*, there are a number of developers supporting GTK+. It might be very difficult for TrollTech to make a profit doing th same thing (hell, it's hard for a company to compete with lots of free software), but that doesn't mean that users should then use TrollTech software.
You're probably right, but in this case at least, the move comes as no surprise. TrollTech has caught flak for years for their licensing policies, and this change should not come as a terrible surprise.
I don't see what the issue is, exactly. DNS data is propagated lazily. The only issues is that you'd have maybe three machines storing the data instead of one.
Unless.torrent files are particularly big -- I happened to have one on my hard drive, which was under 512 bytes.
I'm not sure that there's much point in using DNS to propagate.torrent files -- it seems that USENET or similar would be a better choice, given that they tend to only be useful for a short period of time, that announcements of new torrents is a useful characteristic of a.torrent propagation system, and that archiving torrents is useful -- but I also don't really see the harm in this.
Given the kind of load that nameservers happily handle today when you hit a webpage with a number of entries (especially for those annoying little "badges") (and the nameserver potentially gets twenty or more lookup requests all at once), there can't be a huge processing hit.
There *might* be a storage hit...but suppose there are 10,000 torrent files out there, and each is 1K. That's just 10MB of data, and I doubt anyone is interested in storing all available torrents.
Finally, I suppose that bandwidth might be an issue, but I suspect that given the frequency of DNS lookups and the infrequency of someone needing a new.torrent file, the bandwidth will not be an issue.
I have done plenty of fun things with DNS and run a small DNS server, but I will freely admit that I am not a DNS wizard, and leave it to the folks on NANOG to debate the merits of this.
For my money, though, this is cute and not harmful at all, though it might not be particularly useful.
First, there are a number of core things that people do every day. If someone comes out with a faster, snappier program to do the same thing, I'm likely to prefer said program.
Second, lots of machines run a *lot* of programs today. Sure, my machine is quite capable of running a Freenet instance, even if the thing is a memory hog. But Freenet + Mozilla + GNOME + some stuff in GIMP starts to add up. There is no requirement to have a single application consume all the resources on a system because it can.
Third, there are essentially no trivial GUI programs that I can think of that I use. I do use solitaire, and sure enough, the two implementations that I use are mostly implemented in Scheme and Python. However, OpenOffice is already slow, and would suck very much if it were written in Python. I would not want a bash or a gnumeric written in python. For apps like Calculator, file-renamers, and small front ends -- the sort of thing that are done on Windows in Visual Basic -- I agree.
I know. I was not aware that arts was not considered part of KDE, though -- esound is considered part of GNOME.
You're not seriously claiming xmms as an example of good GTK+ programming are you?
No. I know little of the quality of the GKT+ style in xmms. I've never looked at the source, aside from a brief stint playing with the idea of writing a beat-detection xmms plugin, and GUI code was not high on my list of things to examine.
Sure, the back end is superb, but it looks like crap and no amount of skinning can fix it.
[shrug] I suppose. My idea media player would probably run entirely as a daemon and have "control panels" that could be opened via a pipe to adjust volume and the like.
I should point out that I develop using GTK+ at work and KDE/Qt for pleasure, and Qt is so much nicer to work with it's not even funny.
Hmm. I read a number of John Steinbeck's books years ago for pleasure and enjoyed them. I then read his Grapes of Wrath for a school assignment and couldn't stand it. I've also read Catch-22 under both work and pleasure conditions and found that I liked it much more when reading it for fun.
More modular -- major chunks broken out into the glib and pango components, which can be used entirely independently of gtk. I've used (and written) a number of non-gtk apps that use glib.
Faster -- gtk widgets are drawn more quickly than their qt equivalents, and gtk-based apps start up more quickly than their qt equivalents.
What are you comparing exactly? GNOME vs KDE under?
x86, on my PIII/550 (upgraded recently from a PII/266) Linux box.
It's not like glib doesn't reimplement the STL and has GString.
Sure, but glib isn't written in C++. That functionality isn't present, so glib isn't introducing functionality conflicts with the language that it's written in. If C had a built in socket layer and glib included an incompatible socket layer, then I'd complain bitterly.
Unfortunately, marketing geniuses - the same people who whien about Google's ranking - decided that the system wasn't good enough for them, and started buying domain names without thought for their intended uses.
To lay blame at the proper door, a good chunk of this was due to unscrupulous domain name registrars (mostly Verisign, who was the earliest heavy advocate of this I remember, asking you if you also wanted to buy.net and.org domains when you purchased a d.com).
Slashdot Mirror
Is there a win32 ext2/3 filesystem driver out there anywhere?
Forget that -- there is FAT code in the Linux kernel. More IP that smacks Linux and means that it cannot be distributed (and interoperate with windows, as FAT-based systems were the only major filesystem that both Linux and Windows can read and write out-of-box. Very bad juju.
FWIW, it is *damned* hard to write Windows filesystem drivers -- compare a small Linux filesystem -- RAMFS, at 342 lines of source -- with even a minimal Windows driver. There is an ext2 implementation with read support, though.
Oh, yes. The embedded community uses FAT all over the place. They are going to absolutely go bonkers when this hits the news.
I know that it's not quite what you wanted, but using Coda, which is designed to support disconnected operation (i.e. servers goes away for a while and then comes back) may be an appealing option for you.
(I am a physicist and I have done lots of tests involving 9mm bullets)
I always suspected that you physicists were plotting, up to no good.
Does that make the facts in his post any less relevant or true?
That particular element does not, he's also made plenty of errors in his posts.
However, I do claim that if he wants his posts to be uncriticized, he should remove the false credentials and try for recognition based on the value of the content in them. He has made the choice to impersonate someone else -- he has to live with people criticizing him.
Furthermore, there is social damage caused by impersonating other people to the Slashdot community and propagating false information. This is why people that make errors in their posts frequently see corrections posted. It is the intelligent thing to do for other Slashdot members to resist impersonation becoming accepted.
That really, really sucks.
One of the major selling points of Bungie games is that they have really good co-op.
Sigh...but I can't reasonably expect a Linux version.
:-(
Damn it, I should really sit down and spend a couple weeks trying to figure out what would be involved in some sort of system that would allow binary distribution of (fast) software under Linux that wouldn't break in twenty four months. Stallman may love open source, but the world's always going to have some form of binary-only software...and Linux trails Windows in this arena terribly.
TA was a *fantastic* game.
There were a couple of really neat things it did.
It was one of the first games to leave Blizzard-style micromanagement. The interface is designed so that using it isn't one of the challenges of the game (limited group sizes, queues, etc), but to help you as much as possible (flexible AI toggles per group or unit, easy to queue up masses of units and preassign orders, etc).
It was the first I know of to have really neat sea battles. Infantry were cheap -- you could churn out tons -- but each ship was *expensive*, and specialized. The first time I played a sea battle level, I was enthralled.
It had great explosions and fires.
Battles took place over more realistic ranges -- people didn't shoot the equivalent of twenty feet. The biggest guns could lob rounds from seven or more screens away.
There was no limit on resources. A round didn't come to an end because you exhausted your resources -- you used everything possible, just as intelligently as you could.
There were masses of intelligent auto-build and repair abilities.
And a ton of other things.
Cavedog (TA's publisher) could have gone far, but for two factors: Blizzard, it's main competitor, didn't make as good games but had a phenomenal marketing budget that it used well, and TA's sequel, TA:Kingdoms, really sucked compared to TA.
Incidently, the guy that designed the TA system (where you could tell things to follow things that attacked them, or not etc)...I believe his name was "Tim" something...went on to make some medieval game with the same style interface. It wasn't an RTS, though. I can't remember the name. Fantastic to see that one game designer is interested in making a highly usable interface, not one that you have to fight.
I've read both aritcles and I'm still confused. What the heck is the point and why should I care?
Brief summary (and I'm not informed about this particular spat). LinuxGazette was originally a newsletter that one guy decided to put out to "help people have a just a little more fun" with Linux. He reminded me a lot of Cliff Stoll -- folksy, nice, and very into just helping people out.
As it happened, more and more volunteers started joining, and the LG became one of the early sources for good Linux information and tips.
The folks that publish Linux Journal (a commercial, nonvolunteer publication) started getting involved at some point, paying for an editor and hosting.
Now, the publisher (which has put money and time into Linux Gazette) and at least a large chunk of the volunteers disagree about whether or not to change the format of the gazette.
Frankly, I really don't see the problem. I understand that if you've spent a lot of time working on something, you get upset if you lose it.
Here's my take:
SSC/Linux Journal should get the trademark (legally, if not morally). They registered the thing. It's ugly, but that's life. Not sure whether the fact that it was around for a long time pre-registration is an issue.
SSC/Linux Journal needs to stop modifying copyright information, removing article authors and replacing them with itself. That was a complaint of the forkers. They are quite in the legal right.
The readers deserve to be informed. Putting up information about this on both sites and linking to each other is important.
The readers will go to whatever they want -- if one sucks, they'll leave it.
It's sad that such friendly volunteer work wound up in such a legal tangle. I guess that the moral of the story is: avoid companies unless you're damned sure that they won't screw you over.
When did this fixation that books were somehow 'superior' to visual media first come into vogue? I've seen some very moving movies in my time, and read some awful books.
Sure (for a *very* fun combination, try reading Do Cyborgs Dream of Electric Sheep? and watching Blade Runner). However, the LoTR series is a very, very good series of books. It's a good set of movies, but just good, not very, very good. It's as if someone was a master oil painter and someone made a good watercolor reproduction of one of their works -- the watercolor just isn't a substitute for the original.
Either Mirriam-Webster is in error, or it is referring to a highly unusual and archaic grammar usage.
Keep in mind that Mirriam-Webster may be considered authoritative (well, *I* consider it authoritative, though the really anal-retentive folks will look for the OED) for spelling -- not for grammar. I have always seen English style references state that use of the apostrophe-"s" combination is incorrect. All of the serious resources that Google digs up on short notice (aside from m-w) also back this -- take a look at the MLA style apostrophe guidelines or this linguist's lengthy analysis with an eye on "BOUNDARY MARKER (EXCLUSIVELY)".
Just to ensure that people are aware of the fact -- Samir Gupta is not a PhD, does not work for Nintendo, and is one of Slashdot's more colorful frauds. You can see the beginning of the Samir Gupta hoax on USENET years ago, in early discussions on the Sega Genesis.
However, an amazing number of new people with mod points, impressed with the bogus credentials, frequently mod up his posts.
I'm starting to suspect that the United States will never see conversion.
We have fast enough and cheap enough hardware now that it's feasible (nicer) home connections to stream down much-better-than-TV video over an Internet connection. There are a number of improvements to make in upstream distribution structure, but ultimately, despite the fact that IP currently provides essentially nothing by way of real-time guarantees, my guess is that we'll slowly start seeing more and more Internet-based systems. It just doesn't make sense to have a single purpose dedicated system just for TV.
I suspect that those cheap consumer broadband routers will start having a "smart bandwidth allocation" feature that the ISP will also grok which guarantees real-time delivery (well, over the last and slowest leg of the trip). It wouldn't be a very difficult system to devise -- system on local network allocates bandwidth from router, router talks to upstream system.
A healthy amount of precaching would be important -- this could be an issue in sports, where having a sub-one-minute precache is essential to many hardcore fans. It'd work wonderfully for almost anything else, though.
It'd only really be practical if you put together a massive collection of .torrents (like the suprnova database) in a .zip file and then used BT to download that.
.torrent points to a file. The contents of that file may not be changed without invalidating the .torrent link. Thus, the contents of the file must be known at the time of propagating the .torrent. Thus, the file may only reference older .torrents -- that is, .torrents that were already in existence at the time the file was created. So if the only reference someone has to obtain BitTorrent information is a .torrent file, they may only ever reach .torrents created at or before the creation time of that .torrent.
Good explanation, but it still wouldn't work. (This is an interesting problem that you'll run into in a lot of areas of computer science, though I don't know whether it has a name.) Basically, each
Your solution would reduce load over the existing system, but from the fact that the parent mentioned chicken/egg, I suspect that they want to use BitTorrent as a complete solution.
There are systems that allow providing forward references -- allowing one to reach newer files. These systems require the use of encryption, not just hashing, however. kast and freenet both use public key cryptography to provide exactly this service.
Oh, and if you think you can steal market shares from, let us say, Sun, without them making a fuss, I think you are mistaken too. Last time I checked, Sun is still worth more money than Red Hat...
Last time I checked, Solaris was losing market share rapidly to Linux. Dunno how much of that is to Red Hat Linux, but we can surmise a fair amount.
If Red Hat isn't marketing a UNIX clone, then what's it marketing now? Last time I checked, Linux is a UNIX clone. Sure, it's not SCO UNIX(R)(TM), but it's still UNIX. Sometimes I wonder whether these MBAs really know what the hell they're trying to sell or if they just have a form process to market anything.
No, what he said was exactly right.
"We are making a product foo, which is a clone of bar. Foo competes mostly with bar, and will kill off bar within a decade."
How hard is that to understand?
Weavers are a clone of triscuits, and saying that "triscuits will be dead within the decade, killed by weavers" is an entirely valid statement.
Soon we may wind up with something where the proverbial "Joe Sixpack" pays relatively high fees on his Skype phone he bought at Wal-mart and plugged into the wall, while all the "techies" pay nothing to use their "alternative" VoIP setups.
Remember "audio" CD-Rs? It can happen, little sense as it makes...
MMO is waaay overblown. Simple reason. There just aren't enough players. MMOGs take time to play. There just aren't enough people willing to devote a significant enough chunk of their life to a game. There are some, and they were enough to keep Ultima Online flush with players. There were even enough for Everquest as well. But there are now MMOGs coming out at an incredible clip, which keeps fragmenting the customers base further, and reducing the value of other competitors in the field. MMOGs are flopping left and right, and still publishers frantic for a pirate-resistant subscription model try to start new MMOGs.
How much money has GTK+ made for GNU? How much money has LGPL wxWindows made? How about plain XFree86? I'm not talking about donations from Redhat or SuSE, I'm talking about actual revenue from actual customers. Now ask yourself if that's enough to support even one full time developer?
Yes, but instead of looking at it from a TrollTech point of view, look at it from a user point of view. Regardless of *why*, there are a number of developers supporting GTK+. It might be very difficult for TrollTech to make a profit doing th same thing (hell, it's hard for a company to compete with lots of free software), but that doesn't mean that users should then use TrollTech software.
You're probably right, but in this case at least, the move comes as no surprise. TrollTech has caught flak for years for their licensing policies, and this change should not come as a terrible surprise.
I don't see what the issue is, exactly. DNS data is propagated lazily. The only issues is that you'd have maybe three machines storing the data instead of one.
.torrent files are particularly big -- I happened to have one on my hard drive, which was under 512 bytes.
.torrent files -- it seems that USENET or similar would be a better choice, given that they tend to only be useful for a short period of time, that announcements of new torrents is a useful characteristic of a .torrent propagation system, and that archiving torrents is useful -- but I also don't really see the harm in this.
.torrent file, the bandwidth will not be an issue.
Unless
I'm not sure that there's much point in using DNS to propagate
Given the kind of load that nameservers happily handle today when you hit a webpage with a number of entries (especially for those annoying little "badges") (and the nameserver potentially gets twenty or more lookup requests all at once), there can't be a huge processing hit.
There *might* be a storage hit...but suppose there are 10,000 torrent files out there, and each is 1K. That's just 10MB of data, and I doubt anyone is interested in storing all available torrents.
Finally, I suppose that bandwidth might be an issue, but I suspect that given the frequency of DNS lookups and the infrequency of someone needing a new
I have done plenty of fun things with DNS and run a small DNS server, but I will freely admit that I am not a DNS wizard, and leave it to the folks on NANOG to debate the merits of this.
For my money, though, this is cute and not harmful at all, though it might not be particularly useful.
A couple of reasons I still don't entirely agree.
First, there are a number of core things that people do every day. If someone comes out with a faster, snappier program to do the same thing, I'm likely to prefer said program.
Second, lots of machines run a *lot* of programs today. Sure, my machine is quite capable of running a Freenet instance, even if the thing is a memory hog. But Freenet + Mozilla + GNOME + some stuff in GIMP starts to add up. There is no requirement to have a single application consume all the resources on a system because it can.
Third, there are essentially no trivial GUI programs that I can think of that I use. I do use solitaire, and sure enough, the two implementations that I use are mostly implemented in Scheme and Python. However, OpenOffice is already slow, and would suck very much if it were written in Python. I would not want a bash or a gnumeric written in python. For apps like Calculator, file-renamers, and small front ends -- the sort of thing that are done on Windows in Visual Basic -- I agree.
KDE does not use Glib, Arts does.
I know. I was not aware that arts was not considered part of KDE, though -- esound is considered part of GNOME.
You're not seriously claiming xmms as an example of good GTK+ programming are you?
No. I know little of the quality of the GKT+ style in xmms. I've never looked at the source, aside from a brief stint playing with the idea of writing a beat-detection xmms plugin, and GUI code was not high on my list of things to examine.
Sure, the back end is superb, but it looks like crap and no amount of skinning can fix it.
[shrug] I suppose. My idea media player would probably run entirely as a daemon and have "control panels" that could be opened via a pipe to adjust volume and the like.
I should point out that I develop using GTK+ at work and KDE/Qt for pleasure, and Qt is so much nicer to work with it's not even funny.
Hmm. I read a number of John Steinbeck's books years ago for pleasure and enjoyed them. I then read his Grapes of Wrath for a school assignment and couldn't stand it. I've also read Catch-22 under both work and pleasure conditions and found that I liked it much more when reading it for fun.
How is GTK faster and more modular than Qt?
More modular -- major chunks broken out into the glib and pango components, which can be used entirely independently of gtk. I've used (and written) a number of non-gtk apps that use glib.
Faster -- gtk widgets are drawn more quickly than their qt equivalents, and gtk-based apps start up more quickly than their qt equivalents.
What are you comparing exactly? GNOME vs KDE under?
x86, on my PIII/550 (upgraded recently from a PII/266) Linux box.
It's not like glib doesn't reimplement the STL and has GString.
Sure, but glib isn't written in C++. That functionality isn't present, so glib isn't introducing functionality conflicts with the language that it's written in. If C had a built in socket layer and glib included an incompatible socket layer, then I'd complain bitterly.
Actually, Google ignores META tags, because people try exploiting them, and PageRank style systems work better at judging content.
I'm with you that Google is both in the right and a fantastic company, though.
Unfortunately, marketing geniuses - the same people who whien about Google's ranking - decided that the system wasn't good enough for them, and started buying domain names without thought for their intended uses.
.net and .org domains when you purchased a d.com).
To lay blame at the proper door, a good chunk of this was due to unscrupulous domain name registrars (mostly Verisign, who was the earliest heavy advocate of this I remember, asking you if you also wanted to buy